Does Leading Modality Influence Multisensory Integration in Children with Autism?

Background: Atypical sensory processing is a key component of Autism Spectrum Disorder (ASD) diagnostic criteria (APA, 2013). The process of multisensory integration (MSI) - the ability to efficiently integrate stimuli from multiple sensory modalities - is necessary to experience the world as a coherent whole. Research suggests that disrupted MSI may partially underlie atypical sensory behaviours in ASD (Wallace and Stevenson, 2014). There has, however, been mixed evidence as to whether MSI is altered in individuals with ASD, especially in stimuli void of social content (e.g. Bao et al., 2017). It has been suggested that MSI, in particular tolerance of asynchrony, is largely influenced by leading stimulus modality in typical adults (Cecere et al., 2016). Although this phenomenon has been studied in relation to autistic traits (Stevenson et al., 2017), this has yet to be explored in individuals with ASD or at different periods of development.

Objectives: Assess the contribution of leading modality on tolerance of asynchrony in children with ASD vs typical children at different periods of development.

Methods: Eighty-six participants were separated into ASD child (n=17; age <12), ASD adolescent (n=13; age > 12), TD child (n=30; age <12) and TD adolescent (n=26; age > 12) groups. Participants evaluated the simultaneity of audiovisual stimulus pairs at seven varying stimulus onset asynchronies (SOAs: 50ms, 100ms, 150ms, 200ms, 250ms, 500ms, 750ms) with 7 SOAs presented audio-leading (A-L), 7 SOAs visual-leading (V-L), and one synced (Powers et al., 2009). Participants were asked to respond whether the auditory stimuli (10ms, 1800 Hz pure tone) and visual stimuli (white annulus subtending 9 visual degrees) were presented at the ‘same’ or ‘different’ time.

Results: Proportion of ‘same’ responses were calculated for each participant. A 3-way mixed ANOVA (4 groups x 2 leading modalities x 7 SOAs) revealed an overall significant effect of leading modality (p <0.001), with V-L trials being indicated as synchronous more often than A-L trials, suggesting a greater tolerance for asynchrony for visual-leading stimulus pairs. No significant overall effect of group was found. However, a significant modality by group interaction was found (p <0.001). This effect is explained by a shift in modality sensitivity across age, mediated by group. Whereas only small changes in proportion of ‘same’ responses occurred across age for the ASD group (children; M_A-L =62.43 vs M_V-L =68.25 | adolescents; M_A-L =51.65 vs M_V-L =60.71). Meanwhile, a larger difference in audio- vs visual-leading trial performance was found for the TD group (adolescents; M_A-L =53.00 vs M_V-L =67.45 | children; M_A-L =54.68; M_V-L =61.23).

Conclusions: Asynchrony was tolerated to a greater extent in V-L stimuli for all groups. However, performance was mediated by age group, with a larger developmental jump in visual- vs auditory-leading performance in TD groups compared to ASD groups. Our results provide evidence for atypical development of MSI in individuals with ASD, as defined by tolerance to asynchrony, and provide a platform for further investigation of the developmental trajectory of MSI in this population.